Reconciling Prior Geologic Information with Production Data Using Streamlines: Application to a Giant Middle-Eastern Oil Field

Abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractEffective history matching of real fields requires the resolution of two outstanding problems. First, a conflict may exist between the production data and the existing geological model built solely from static information. In resolving this problem one must relate the inherent multi-scale nature of production data to petrophysical properties of the reservoir at the proper scale. Second, during model updates, geological consistency must be maintained by honoring the prior geologic information. The geoscientist has to choose what prior information is well know (hence fixed), and what is uncertain (hence modifiable).In many instances, the type of geostatistical algorithm is fixed, while key prior geostatistical parameters should be perturbed (e.g. facies proportions, petrophysical properties trends, variogram parameters, and random seed).We propose a new methodology that addresses these problems. First, streamlines are used to relate the production data to petrophysical properties at multiple scales. A combination of geostatistical tools (locally varying mean and probability perturbation method) are then used to jointly map multi-scale corrections back to the geological model through changes of the prior geostatistical parameters. The mapping reconciles the fixed prior geologic information with the production data. The geological model is then explicitly recreated by re-running the geostatistical simulation. This approach differs from other history matching techniques where the petrophysical properties of each grid block are modified directly. While a successful history match may be obtained, the resulting model may be inconsistent with important prior information, hence retaining little predictive power.The methodology is demonstrated by applying it to history matching a giant Saudi Arabian carbonate oil reservoir. This reservoir has over 500 producers and over 50 years of historical data with dramatically changing field conditions. In the past, several attempts were made to manually history matching this reservoir. The process was found to be extremely time consuming, involving dramatic local permeability changes which are often not supported by geological data. By applying the new approach, rather than correcting permeability manually, the corrections supplied by the streamlines were used to constrain the geostatistical algorithms, thereby ensuring a consistent geological scenario at every iteration. * 1 c . The new LVM is passed back to the geostatistical algorithm, and the permeability field is rebuilt with a new trend accounting for the water breakthrough data.

Polymer-Flood Modeling Using Streamlines

“…The speed of SL simulation and novel solution data such as well allocation factors and well drainage/irrigation zones have made SLs an important, complementary approach to traditional simulation approaches in reservoir engineering workflows such as sensitivity runs, evaluating upscaling algorithms (Nair and Al-Maraghi 2006;Samier et al 2002), flow-field visualization, full-field simulations, ranking and uncertainty quantification, surveillance (Batycky et al 2008;Grinestaff 1999), flood management (Ibrahim et al 2007;Thiele and Batycky 2006), and history matching (Batycky et al 2007;Fenwick et al 2005;Milliken et al 2001), just to mention a few. The modeling of complex systems as might be the case in compositional simulation (Gerritsen et al 2007) and fractured reservoirs (Di Donato et al 2003), where the primary displacement mechanism is the subtle interaction of local displacement efficiency and interwell channeling, has also been extended to SL simulation.…”

Streamline Simulation with Compressibility and Spatial and Dynamic Variation in Oil Composition

“…The updated masses and saturations at the end of the time step are mapped back to the original grid. Nowadays streamline simulation is used in a range of applications that includes history matching (Emanuel and Milliken 1998;Maschio and Schiozer 2004;Cheng et al 2007;Stenerud et al 2008;Gross et al 2004;Fenwick et al 2005), quantification of uncertainties (Christie et al 2002;Ligero et al 2003), flood surveillance (Batycky et al 2008), CO 2 storage (Qi et al 2007), compositional simulation (Thiele et al 1997;Osako and Datta-Gupta 2007), and dual porosity modeling of fractured reservoirs (Di Donato et al 2004;Kozlova et al, 2006).…”

Assisted history matching using pattern recognition technology

“…Later Agarwal and Blunt (2003) extended the work of Wang and Kovscek to a real field case in the North Sea. Fenwick et al (2005) used streamline simulation coupled with a combination of geo-statistical tools to history match a giant Middle Eastern oil field. They just changed the permeability to achieve the match.…”